We propose to continue one of the most complete prospective longitudinal studies of growth and development yet undertaken of children at high risk of obesity. The study of 72 children now 9 to 10 years of age, selected on the basis on maternal overweight or leanness, achieved its initial goal, disconfirming the widely held belief that a low total energy expenditure predicted body size and composition as early as one year of age. It has since found that, at 2 years of age, 33 children of overweight mothers did not differ from 37 children of lean mothers in any aspect of body composition. By 4 years of age, high risk children had greater weight, body mass index (BMI), skinfold thickness and lean body mass. By 6 years of age they also had significantly greater levels (25% vs. 19%) (6.7 kg to 4.8) of body fat. Ten had exceeded the 85th percentile for BMI (odds ratio 15.7). We now propose to study the development of these subjects into adolescence focusing on behavioral determinants of body size: energy intake and physical activity. We will first evaluate the powerful influence of familial vulnerability on growth and development. Next, we will test the influence of food intake and physical activity on body composition, and determine the extent to which the influence of familial vulnerability is mediated by these behaviors. Energy intake will be assessed by food records, recalls, laboratory protocols and 2 questionnaires. Physical activity will be assessed with activity monitors and with monitors of television and computer use and by recalls. These efforts will be enhanced by the remarkable sample maintenance, 92% at 10 years (72 out of the original 78 subjects) and no subject has been lost during the past 5 years. Parents report that they will continue to the end of the grant period. These subjects will make it possible to assess the effects of behavioral factors over time in a cohort that has been identified as being at high risk for obesity and thereby elucidate critical gene-environmental interactions.